On 2006-01-16 15:36, alexandru_mg3 wrote:
> But in the initial position Latin gWHr- > gr-
No. This is the development of unaspirated *gWr-. The combination *gWHr-
was rare in PIE, and I don't know of any certain examples of its
development in Latin. *gWHr- > gr- is a priori unlikely, given that
*g(^)Hr- developed into *hr- > r-. Judging from the behaviour of other
sounds from the natural classes to which *gWH- belongs, I'd expect
*gWHr- > Lat. fr- (before a vowel, *g(^)H- > *x- > h-, *gWH- > *xW- >
f-). Medially between vowels, *gWH- > Lat. v. The development of medial
*-gWHr- > *-Br- > -br- (parallel to *dHr, *sr and *bHr, all of which
yield initial fr- and medial -br-) is both expected and supported by
examples. Note, additionally, *negHWr- > nefr-, nebr- in other Italic
dialects, and Umbr. vufru 'votive' < *(h)wegWH-ro-.
> Next, this medial Lat. *-gWHr- > Lat *br is based Only on 2 examples
>
> First: '*dHegWH-ri- > febris' and another one that I don't remember
> now but I will check in the evening at home.
You probably mean the <e:brius, so:brius> set.
> P>S> By the way (if the above etymology is the right one) see the
> function of -r- in Febris if is similar-to *dHegWH- or not (and if
> febris has or not 'a different nature' than *dHegWH-
A trivial observation. *dHegWH- is a _verb_ root, so of course the
semantic relationship with a related noun via a chain of derivatives is
indirect. A fever is often described as a "burning sensation".
Piot